Re: [Atoca] Requirement D2: "Large Audience"

["mark.wood" <[e-mail]>]

Yes, I stand corrected, Thanks Richard.



-----Original Message-----
From: Richard L. Barnes [mailto:[e-mail]] 
Sent: Monday, January 17, 2011 3:41 PM
To: Brian Rosen
Cc: <[e-mail]>; [e-mail]
Subject: Re: [Atoca] Requirement D2: "Large Audience"

Just one brief comment: Mark's statement that "it's only possible to have
any kind of delivery reports regarding specific terminals on a one-to-one
kind of unicast delivery" is not necessarily true.  The availability of
delivery reports depends entirely on the capabilities of the terminals, and
not the intervening access network.  In particular, terminals can send back
delivery reports regardless of whether they are sending or receiving over
broadcast or unicast.   

This pattern could actually be quite efficient in certain layer-2 networks
that have broadcast downstream, unicast upstream (e.g., PON, DSL, maybe
Cable IIRC).  A notification to an entire subnet takes one message, and all
acknowledgements can be received in parallel.

--Richard



On Jan 17, 2011, at 9:44 AM, Brian Rosen wrote:

> It may be, but I'd like to explore this a bit anyway.
> 
> Millions of messages (acknowledgements) is a scale we can deal with today.
Hundreds of millions is probably beyond what we can deal with in a response
to a very large alert.
> 
> Most systems consist of several smaller subsystems.  The purpose of an
acknowledgement is to make sure everyone got the message.  If the subsystem
can determine that every one of its clients got it, it can report that up
the line.  It can save missed acks for later analysis, or if there are few
enough of them, report them up.
> 
> This means messages national scale which have small effectivity times
can't reasonably ask for message acknowledgement.  Anything smaller than
that probably can.
> 
> Since most alerts really don't involve hundreds of millions of
notifications, most alerts probably can ask for them.
> 
> If your delivery mechanism is multicast, the multicast mechanism itself
doesn't track who gets the alert in any way we can use.  That implies
something else is tracking who gets the alert, a complication that could
loom large.  Some systems do know who gets the alert (sometimes because it
knows who it is connected to, and all of them get the alert).  Certainly,
anything with a subscription has the characteristic that the sender knows
who all the recipients are.  
> 
> It's VERY valuable to know that every entity that should get the alert got
it.  The only other mechanism we have is some repeating of the sending in
the hopes that everyone got it.  In some cases you may have more than one
"path" to the same recipient.  That might be multiple devices, multiple
services, or multiple logical or physical connections.  You may try one
first, and if that doesn't get an ack, try another.  Although we often think
of this mechanism as needing no more than seconds to deploy, in fact many
alerts would be fine with a few minutes, and trying some things sequentially
may make sense.
> 
> So, yes, probably a Tsunami alert to all of East Asia can't ask for
acknowledgements.  An "Amber Alert" (possible abducted child) to a county
might very well.  Certainly, a snow emergency closing to the parents of an
elementary school could.
> 
> Brian
> 
> 
> On Jan 17, 2011, at 6:03 AM, <[e-mail]>
<[e-mail]> wrote:
> 
>> Hello chaps,
>> 
>> Yes, it's only possible to have any kind of delivery reports 
>> regarding specific terminals on a one-to-one kind of unicast 
>> delivery. The Australians were not considering Cell Broadcast or any 
>> other kind of multicast technology as they have selected a SMS based 
>> delivery system, so this is probably what they had in mind. I noticed 
>> that atoca language about this is reserved mostly for the 
>> 'subscription push' kind of distribution.  We need to take care not 
>> to make it look like a mandated requirement for broadcast/multicast
bearers.
>> 
>> IMHO, the other special problems with multicast,  mean that it needs 
>> different and special requirements and cannot be generalized with 
>> other 'push' or 'pull'  technologies quite so easily, tempting though 
>> that is. I think this problem happens because most IT professionals 
>> are familiar with TCP and UDP but rarely use multicast so are 
>> unfamiliar with its very different attributes.
>> 
>> Many vendors offer systems which do have a positive indication of who 
>> got what,  and this is a good thing for applications of up to about 
>> 10,000 receivers. Netherlands studies showed that by the time you 
>> scale up to millions,  the mass scale of all this becomes a problem, 
>> so large scale systems will have to reluctantly abandon the notion of  
>> individual confirmation.
>> 
>> Besides, the police authority of a large city don't have the time, 
>> during an emergency, to check out millions of replies nor take any 
>> action if someone does not reply (they are too busy). On the other 
>> hand a smaller community may value such a facility, so it seems that 
>> they authority will need to choose whatever is appropriate for the
circumstances at their discretion.
>> There is no 'one size fits all' solution so we will have to blend 
>> many bearers in any real situation.  For example the Norwegians take 
>> the opposite view.
>> 
>> However, it is possible to confirm that the multicast was 
>> transmitted, as a network of feedback receivers is envisaged in 
>> conjunction with an outer loop reporter system, which would report of the
message was transmitted on air.
>> If it did not, then an alarm would be generated by the originating 
>> aggregator for the log file. So the positive indication of the fact 
>> of the transmission over the air is satisfied even though individual 
>> delivery is not.
>> 
>> In summary, IMHO,  Broadcast/multicast bearers will need special 
>> treatment as they don't neatly conform to the classic notions of 'push'
or 'pull'.
>> 
>> Warm regards, Mark  Wood.
>> 
>> 
>> 
>> 
>> _______________________________________________
>> earlywarning mailing list
>> [e-mail]
>> https://www.ietf.org/mailman/listinfo/earlywarning
> 
> _______________________________________________
> earlywarning mailing list
> [e-mail]
> https://www.ietf.org/mailman/listinfo/earlywarning